FINAL REPORT January 31,2004 MICROGRAVITY STUDIES OF LIQUID-LIQUID PHASE TRANSITIONS IN ALUMINA-YTTRIA MELTS Flight Definition, Materials Science Subdiscipline: Ceramics and Glasses NASA Contract Number NAS8-98092 April 1998 to January 2004 Principal Investigator: Dr. Richard Weber Co-Investigator: Dr. Paul Nordine Containerless Research, Inc. Evanston, IL 6020 1-3 149 Phone: 8471467-2678, Fax 8471467-2679 email: [email protected] COTR: Buddy Guynes George C. Marshall Space Flight Center Marshall Space Flight Center, AL 35812 Proiect Summary: The scientific objective of this research is to increase the fundamental knowledge base for liquid- phase processing of technologically important oxide materials. The experimental objective is to define conditions and hardware requirements for microgravity flight experiments to test and expand the experimental hypotheses that: 1. Liquid phase transitions can occur in undercooled melts by a diffusionless process. 2. Onset of the liquid phase transition is accompanied by a large change in the temperature dependence of melt viscosity. Experiments on undercooled YAG (Y3A15012)- and rare earth oxide aluminate composition liquids demonstrated a large departure from an Arrhenian temperature dependence of viscosity. Liquid YAG is nearly inviscid at its 2240 K melting point. Glass fibers were pulled from melts undercooled by ca. 600 K indicating that the viscosity is on the order of 100 Pans (1000 Poise) at 1600 K. This value of viscosity is 500 times greater than that obtained by extrapolation of data for temperatures above the melting point of YAG. These results show that the liquids are extremely fragile and that the onset of the highly non-Arrhenian viscosity-temperature relationship occurs at a temperature considerably below the equilibrium melting point of the solid phases. Further results on undercooled alumina-yttria melts containing 23-42 mole % yttrium oxide indicate that a congruent liquid-liquid phase transition occurs in the undercooled liquids. The rates of transition are inconsistent with a diffusion-limited process. This research is directed to investigation of the scientifically interesting phenomena of polyamorphism and fragility in undercooled rare earth oxide aluminum oxide liquids. The results bear on the technologically important problem of producing high value rare earth-based optical materials. Approach: The research is focused on the study of undercooled pseudo-binary alumina-yttria melts. Containerless techniques are being used to access non-equilibrium liquids and allow precise control of melt chemistry at high temperatures. Although the oxides of interest are extremely stable and exhibit only small oxygen composition changes, some properties of melts are very sensitive to the ambient oxygen pressure which must be controlled in order to investigate subtle effects in liquid oxides. Properties under investigation are the onset and kinetics of phase transitions in the liquid and the viscosity of the undercooled liquid as a function of the temperature. Earth-based research is being performed under conditions that meet some of the experimental constraints to provide the 1 design basis for microgravity experiments. Containerless experiments using aero-acoustic (AAL), aerodynamic (CNL) and electrostatic (ESL) levitation techniques achieve precise control of melt chemistry, access highly undercooled melts and glassy states, and control the effects of interaction with container walls on the evolution of morphology. Microgravity experiments will obtain the control of density-driven segregation, and especially of fluid flow required to test the experimental hypotheses and enable measurements of melt viscosity over the range of experimental conditions of interest. Ground-based containerless experiments employ CRI’s aero-acoustic and aerodynamic levitation facilities with C02 laser beam heating to investigate the effects of processing conditions on the compositions of interest. In these experiments, the melts are highly stirred by the aerodynamic and acoustic levitation forces and equilibrate rapidly with the process atmosphere. The experiments are performed in atmospheres with a controlled p(02), which (i) preserves the melt stoichiometry and avoids partial reduction of the oxides, and (ii) achieves conditions under which the liquid phase separation and glass formation occur in the YAG-composition liquid. Ground-based research using the ESL facility at NASA Marshall Space Flight Center (MSFC) is investigating the requirements for ESL experiments on oxides and defining boundary conditions for the flight experiments. The requirements for flight experiments are described in the Draft Science Requirements Document. Detailed designs for the space-based microgravity experiments will be developed in collaboration with NASA scientists and engineers. This work may include low gravity experiments aboard parabolic aircraft flights if the ESL becomes available for this purpose. Activities and Results: The major activities during this project were to: 1. Investigation of phase transitions in selected YA-composition materials. 2. Sample preparation and planning for ESL experiments at MSFC. 3. Publication of results. 4. Complete an SCR. 5. Issue press release. Analysis of Single and Two-phase Glasses: Here we briefly present and discuss investigation of single- and two-phase glass materials formed from alumina-yttria compositions. More detailed reports are provided in previous NASA reports, the NASA Taskbook and our published and in-preparation papers. SEM images of sectioned samples of YAG- and Y203 + 76 mole % A1203 (Y76A)-composition glasses are presented in Fig. 1. The YAG-composition glass is clearly 2-phase containing spheroidal precipitates that have the same chemical composition as the matrix. This result is consistent with the interpretation that the liquid undergoes a partial phase transformation prior to 2 vitrification. In more recent work we have concentrated on investigation of the Y76A materials that are at the limit of the glass forming range in the alumina-yttria system. As shown in Fig. 1. the Y76A-composition materials form a single phase glass core surrounded by a thin dendritic crust. SEM analyses performed by Dr. Greg Jerman at NASA MSFC indicate that the core has the same chemical composition as the batch of material used to make the samples. The crust consists of alumina dendrites and a Y-rich phase. Differential Scanning Calorimetry (DSC), see Fig. 2, shows that the Y76A glass exhibits a glass transition at approximately 1150 K followed by two exothermic events. The first exotherm releases about 40 % and the second about 50 % of the total heat released. Some additional heat is released during high temperature thermal events associated with densification of the crystalline phases. Fig. 1: Top left, SEM image of Y62.5A (YAG-composition) glass spheroid showing ‘droplet’ phase in amorphous matrix, top right, close-up image of the droplets in the glass matrix. Bottom left, SEM image of Y76A glass spheroid, showing crystalline surface layer and amorphous core, and bottom right, close-up image of the crystalline surface layer. 3 700 900 1100 1300 1500 200 150 100 50 0 700 900 I loo 1300 1500 700 900 I 100 I300 I500 Temperature (K) Fig. 2a-c: Differential scanning calorimetry (DSC) traces for (a) Y62.5A, (b) Y71.5A, and (c) Y76A, showing glass transition and exotherms. Insets show exotherms in their entirety; note the double exotherm for Y76A. Temperatures marked T2-T5 correspond to the temperatures at which samples were recovered for further analysis. (Tl, not marked, corresponds to the as-made glass.) 4 X-ray diffraction analysis indicated that the materials are amorphous before and after the first exothermic event. The X-ray diffraction spectra show a few small peaks that could not be indexed, and were attributed to traces of crystals present in the as-made glasses. Nanocrystals of YAG (size - 18 nm) are formed in the second exothermic event, and grow to - 48 nm after heating to T5 - 1600K. Density measurements showed a 10% increase in density, from 3.8 to 4.2 g/cm3, after heating the as-synthesized material to T2. To further understand the nature of the effect that produces such a large evolution of heat during the first exotherm, heat treated samples were studied by high-field 27AlNMR, neutron diffraction and transmission electron microscopy, results are shown in Figs. 3,4 and 5 respectively. Fig. 3: High field 27AlMAS NMR spectra for as-made (Tl) and ‘transformed’ (T2) Y76A glass. T1 and T2 correspond to Y76A samples as illustrated in Figure 2c. Note the decrease in 4-coordinate and increase in 6-coordinate A1 ions as a result of the structural rearrangements associated with the first exothermic event (T2) in Figure 2c. 50 -50 Chemical Shift (ppm) 5 2 AY76: as made glass 1.5 1 9 r/o 0.5 0 U 1 I I 5 10 15 20 25 AY76: ‘transformed’ glass n 9 rn ’ 5 10 15 20 25 33 Momentum Transfer: Q (A-’) Fig. 4. Neutron diffraction spectra for the as-made and ‘transformed’ Y76A glasses. The peaks for the ‘transformed’ Y76A glass were indexed to cubic gamma-Al203. 27AlNMR reveals that heat treated glasses contains a much increased population of 6 coordinate aluminum ions compared to the as-made material. Neutron diffraction shows weak, broad Bragg peaks consistent with a nanophase component, indexing of the pattern indicates that the crystalline phases is metastable gamma-aluminum oxide. High resolution TEM imaging shows no evidence of crystals in the as-made glass core. After heating to T2, TEM shows a uniform distribution of nanoclusters of gamma-aluminum oxide in an amorphous matrix. If one assumes that the quantity of gamma alumina does not exceed that required to convert the matrix phase to the YAG composition, then the nanocrystalline gamma alumina would comprise up to 36 mol% of the sample on an M203 basis. Since (i) the transformed material density is - 4.2 g/cm3 and (ii) the density of gamma alumina and as-synthesized YAG-composition glass are both - 4 g/cm3, it appears that a polyamorphic transformation of the matrix glass phase occurs together with the precipitation of gamma-alumina during the first exotherm.